JDrop + d3.js = user data driven visualizations?

[Johan Sundström]

I have been playing with http://mbostock.github.com/d3/ all day, and it occurred to me that it would be great fun giving anybody on the web the opportunity to reuse whatever data visualizations I come up with. All the graphs on the above url are populated from either random data or a json (or csv) file at the moment, so I'd like to craft versions of one or a few of them that just become parametrized on the json url to feed from data supplied by visitors.

Are the jdrop.org/export urls supposed to export plain JSONP, and would they work only for logged-in users or for public use on the web too? I haven't been able to get more than 0-length http responses from them with the links in the public section, and am trying to figure out what the constraints and intent are, as that would inform my app design.

--
/ Johan Sundström, http://ecmanaut.blogspot.com/

[Johan Sundström]

Let me try to respond to that myself, in the interest of sharing some
details on JDrop that aren't tractably findable on the site itself, as
best I can from what experiments and peeking I did:

> Are the jdrop.org/export urls supposed to export plain JSONP

No.

> and would they work only for logged-in users

Yes. All JDrop data consumers need to be logged in (via their Google
or Twitter account).

> or for public use on the web too?

No; only logged-in users get JSON from the export page, whether marked
public or not. (Public data might be readable by logged-in users that
did not save the data though, for all I know; I did not look into it
deeper.)

And, for the purposes of what I was trying to do, github's gists are a
better match, I realized on discovering Mike Bostock's (d3 author)
http://bl.ocks.org/ site doing essentially what I was thinking of,
except differently.

--
/ Johan Sundström,http://ecmanaut.blogspot.com/